Ingestion
Connectors to tills, management systems, delivery channels, banks and staff tools. Data arrives every day with quality checks that stop a dirty row before it enters the model.
- Tills
- Management systems
- Delivery
- Banks
- Clock-ins
Data and operations platform
All your data. One platform.
AtomicBI collects every operational, accounting and commercial flow across the network, normalises it into a single data model and returns it as a decision. On web, iPhone and Android, with the same numbers for everyone.
Some of the brands working with AtomicBI
The platform in numbers
Operating figures, kept up to date on the production footprint.
50+
Functional modules in production
100+
Stores connected to the platform
11
Data schemas integrated into a single model
3
Native clients, web, iOS and Android
How it works
From raw data to a decision, in three steps
AtomicBI is a business intelligence and operations platform for multi-site networks. It unifies revenue, accounting, staff, inventory and sales channels into a single data model, and delivers the result to web, iOS and Android with permissions by role, by page and by store.
Normalisation
A single relational data model. Master data aligned, management codes reconciled across different systems, time zones and currencies standardised, history consolidated on fiscal closings rather than open orders.
- PostgreSQL
- Single master data
- Reconciliation
- History
Distribution
Dashboards, scheduled reports, notifications and an AI assistant. Each user sees only the stores and pages their role allows, with the same numbers on every device.
- Web
- iOS
- Android
- Notifications
Modules
Every business area, in the same place
Fifty active modules covering the full chain, from the first receipt to the closing of the accounts.
Business intelligence
Real time dashboards, configurable indicators, period comparisons and interactive charts for every level of the organisation.
Structured reporting
Weekly, monthly and store manager reports, plus a custom report generator with export to PDF, Excel and Word.
Accounting and reconciliation
Revenue control, fiscal takings, cash book, reconciliation of deposits and invoices, payment schedule and cash balancing.
Store operations
Daily closings with automatic formulas, operational checklists, area control, a deposits module and cash tracking.
Advanced analytics
Product analysis, menu engineering, sales by time slot, channel yield and comparison across comparable stores.
Staff
Clock-ins, employee records, shift analysis, labour cost on sales and productivity by site.
Inventory and supply
Bills of materials, delivery notes, historical prices, stock levels, supplier orders and analysis across the whole supply chain.
Food cost and margins
Incidence of purchases on sales, median by brand, deviations above threshold and cost reconstruction per unit of measure.
Marketing and online presence
Website performance, campaigns sent, map analysis, footfall times and collection of market signals.
AI Assistant
A chat that queries the database in natural language, answers with the company numbers and always cites the source of the data.
Granular permissions
Access by page, by store and by role. Reusable permission templates and a dedicated administration panel.
Alerts and deadlines
Push notifications, automatic emails on discrepancies and out of threshold indicators, a schedule and planned jobs that run on their own.
The product
Real screenshots of the platform
The same views managers use every day, from head office to the individual store.
Food cost calculated from documents, not from estimates
The incidence of purchases comes from the delivery notes already in the management system, not from a manual upload. The platform derives the cost per unit of measure from real prices and compares each store with its brand median.
- Median by brand rather than network average, so one anomalous store does not shift the benchmark.
- Deviations above threshold flagged by product category.
- Automatic assistant commentary on the table, without writing any query.
Closings and checklists that check themselves
Every evening the platform collects the cash closing and operational checklist from every store, calculates the differences and highlights only what needs a look. The area manager opens the list and sees the exceptions straight away.
- Balancing formulas applied identically across the whole network.
- Completion from a phone too, with the same questions as the web.
- History available and comparable across periods.
An assistant that reads the data, not one that invents it
The assistant queries the database in read only mode and answers citing the period and the source. It has no write access, does not go beyond the stores the user is allowed to see, and does not answer when the data is not there.
- Read only access, with control over executable instructions.
- Respect for the permissions of the user asking the question.
- An answer with period, source and verifiable numbers.
Multi platform
The same numbers, on every device
The iOS and Android applications are not a web page inside a frame. They are native clients, written in the language of the platform, that talk to the same backend.
Web
The complete environment, built for management and for analysis work on a large screen.
iPhone
A native application for work in the field, from the evening checklist to checking revenue.
Android
The same menu structure and the same modules as the iOS version, aligned view by view.
Security
Company data stays where it should
The platform handles revenue, contracts and staff data. The architecture is built to close access by default and open it only where needed.
Role based access
Centralised authentication, permissions by page and by store, reusable role templates and immediate revocation.
Private network
The database cannot be reached from the internet. Services contact it on a private address through a dedicated network.
Managed secrets
No credentials in the code. Service passwords live in a dedicated store and are rotated when needed.
Traceability
Every extraction leaves a trace. Ingestion anomalies are logged and flagged instead of being ignored silently.
Frequently asked questions
The short answers
The questions that come up most often about what AtomicBI does and how it is used.
AtomicBI is a business intelligence and operations platform for networks with many stores. It unifies revenue, accounting, staff, inventory and sales channels into a single data model and delivers them to web, iOS and Android with permissions by role, by page and by store.
The platform is developed by Investfood SRL and runs on its restaurant network, more than 100 stores spread across seven brands. The modules are written to be sector independent, so the same structure applies to any network managing many sites with accounting and operational data to consolidate.
On three clients. The web covers the complete environment for analysis on a large screen. The iPhone and Android applications are native, not wrapped web pages, and stay aligned with the web view by view. The data is the same everywhere because it comes from the same backend.
Permission is assigned on three levels, the page, the store and the role. Reusable permission templates exist to create a new profile by copying that of a colleague with the same job. A user with no assigned pages sees nothing, not everything.
With tills and store management systems, delivery channels, inventory and supply tools, clock-in systems and banking flows. The connectors run every day and write to a single relational data model after passing the quality checks.
On managed cloud infrastructure, with a relational database reachable only from a private network and a dedicated store for service credentials. The database does not expose a public address and every service that queries it goes through a private network.
Access your workspace
Your credentials open an environment built on your role, your permissions and the stores assigned to your profile.